Friday, April 09, 2010

Krugman on Climate Economics

Paul Krugman has a long, beautiful article on the economics of the climate change problem in this weekend's New York Times Magazine, and you really ought to read it. (Really.) He has a wonderful way of cutting through all the crap that is out there and getting to the heart of the matter, and explaining his ideas concisely and forcefully.

For example, here is his summary of the science:
But before we get to the economics, it’s worth establishing three things about the state of the scientific debate. The first is that the planet is indeed warming. Weather fluctuates, and as a consequence it’s easy enough to point to an unusually warm year in the recent past, note that it’s cooler now and claim, “See, the planet is getting cooler, not warmer!” But if you look at the evidence the right way — taking averages over periods long enough to smooth out the fluctuations — the upward trend is unmistakable: each successive decade since the 1970s has been warmer than the one before.

Second, climate models predicted this well in advance, even getting the magnitude of the temperature rise roughly right. While it’s relatively easy to cook up an analysis that matches known data, it is much harder to create a model that accurately forecasts the future. So the fact that climate modelers more than 20 years ago successfully predicted the subsequent global warming gives them enormous credibility.

Yet that’s not the conclusion you might draw from the many media reports that have focused on matters like hacked e-mail and climate scientists’ talking about a “trick” to “hide” an anomalous decline in one data series or expressing their wish to see papers by climate skeptics kept out of research reviews. The truth, however, is that the supposed scandals evaporate on closer examination, revealing only that climate researchers are human beings, too. Yes, scientists try to make their results stand out, but no data were suppressed. Yes, scientists dislike it when work that they think deliberately obfuscates the issues gets published. What else is new? Nothing suggests that there should not continue to be strong support for climate research.

And this brings me to my third point: models based on this research indicate that if we continue adding greenhouse gases to the atmosphere as we have, we will eventually face drastic changes in the climate. Let’s be clear. We’re not talking about a few more hot days in the summer and a bit less snow in the winter; we’re talking about massively disruptive events, like the transformation of the Southwestern United States into a permanent dust bowl over the next few decades.

Now, despite the high credibility of climate modelers, there is still tremendous uncertainty in their long-term forecasts. But as we will see shortly, uncertainty makes the case for action stronger, not weaker. So climate change demands action.

Who has ever said this better in as many words?

In terms of the economics, Krugman favors a cap-and-trade system augmented with strong restrictions on the usage of coal. He's not in favor of a carbon tax because he thinks it offers too little incentive to individuals -- but I wish he had considered Hansen's idea of a "green check" returned to individuals and households every month (quarterly would probably be better). Personally I think that would be a real winner -- Hansen says most American's (and especially the right ones) would receive more in green checks than they would pay in carbon taxes.

The other major point Krugman makes is that regulating a resource market to control temperature is no different than a market's response to a dwindling supply of that resource.

Plus, let's not forget that the conservative's method choice to control acid rain was a cap-and-trade on SO2. There's little basis for objecting to its as a solution to the CO2(e) problem. (Hence their need to deny reality.)

Anyway, be sure to read the entire article.

12 comments:

rhhardin said...

the upward trend is unmistakable: each successive decade since the 1970s has been warmer than the one before.

You can't tell a cycle from a trend with only a little data. A cycle can't be anthrogenic. The earth has had lots of cycles. What Krugman has to prove is that it's not a cycle.

Second, climate models predicted this well in advance, even getting the magnitude of the temperature rise roughly right.

The surviving climate models predicted it.

Now you have to estimate how many climate models there were, and show that it's not just the fraction you'd expect that got it right.

One coin in a thousand follows ten random calls, proving not that it's an exceptional coin but that you started with a thousand coins.

I can't stress enough what garbage computer models are. Maybe you need more computer modelling experience.

David Appell said...

> I can't stress enough what garbage > computer models are.

Not true, but do you have a better idea for determining future climate?

(I know last time you said to analytically solve the PDE system, but you weren't convincing that that could be done.)

> The surviving climate models
> predicted it.

Of course, that's how it always is in science (and in life), which is why they no longer teach about phlogiston in thermodynamics classes or first ask what it says when writing scientific papers. Or should they?

Dano said...

Poor Ron thrashes about for any kind of play to negate the paradigm that negates his ideology. To no avail.

Ah, well. Whaddyagonnado? Either live your life in denial or wake the F up. Either way, no one cares about those in denial anyways. Non-functioning non-members of the commonweal.

[/harsh belittlement of hopeless psychological condition]

Best,

D

rhhardin said...

It's important, if you're going to use prediction as a measure of validity, that the number of competing models be small compared to the degrees of freedom in the prediction.

Otherwise you wind up with a successful surviving model regardless what you're putting in them.

As to models that might be valid, you can't do them. The PDEs are impossible.

Which is the primary way to know that the models are garbage.

Krugman at least acknowledges some connection between curve-fitting and invalidity, by the way. Possibly some econometrics experience taught him something, back when people liked to model the economy.

Dano said...

Incoherent polysyllaby as authoritative rhetoric - who knew such tactics existed?!?

Best,

D

rhhardin said...

@Dano,

If you put all possibilities into a thousand envelopes, one of the envelopes will successfully predict the outcome of ten coin tosses.

It that a particularly smart envelope?

No, it just was part of a covering of all possible outcomes.

Similarly with predictions from too many climate models.

The principle comes from asking how many times you'll be wrong if you take a perfect match as something other than chance when it's in fact chance.

If you've covered all the field, then you'll be wrong 100% of the time.

If you've covered half the field, then you'll be wrong half the time.

What you want for a good indication of validity is complicated outcomes and few candidates.

That is, many degrees of freedom in the prediction and few attempts at predicting.

Dano said...

Wow. You've just proven yourself a genius worthy of the name Galileo!

Destroying - here, today - the work of decades of climate science with a few deft clicks on your keyboard. When will you publish your comment, this blockbuster of a work, laying waste to a discipline? Let us know.

Best,

D

David Appell said...

> It's important, if you're going
> to use prediction as a measure of
> validity, that the number of
> competing models be small compared > to the degrees of freedom in the
> prediction.

Which is certainly true, and always has been. There are about 20 GCMs taken seriously, and that's larger than in past decades.

David Appell said...

> As to models that might be valid,
> you can't do them. The PDEs are
> impossible.

You still haven't told us how else to estimate future climate. The basic physics of this situation are too clear and too worrisome to just say "it can't be done." Too much depends on the answers.

rhhardin said...

It can't be done. Sorry.

rhhardin said...

As to what's worrisome, I'd think that would depend on susceptibility to worry, to use a neutral term.

Dano said...

It can't be done. Sorry.

Lolz.

Best,

D